What is Known About Cow's Milk Protein Allergy in Preterm Infants?

Abstract

Background:

Cow's milk protein allergy (CMPA) is well described in term infants, as opposed to preterm infants. In preterm infants, CMPA shares many gastrointestinal symptoms with necrotizing enterocolitis (NEC).

Objectives:

To evaluate the presentation of CMPA in preterm infants and to investigate the different diagnostic and therapeutic options.

Materials and Methods:

We searched for the relevant literature using the medical databases PubMed, Web of Science, and the Cochrane Library. We performed a post hoc analysis on the 25 case reports included in this study.

Results:

Literature was scarce and heterogeneous. The majority of preterm infants with CMPA were exposed to bovine-based milk proteins before the development of symptoms. The most common clinical manifestations were bloody stools, vomiting, and abdominal distension. Of the 25 cases, only 7 (28%) retained human milk in their diet after diagnosis. In the larger studies, no study has human milk as primary feeding choice after diagnosis.

Conclusions:

Preterm infants exposed to a type of cow's milk-based formula in their first days of life have a higher risk of developing CMPA. Most of the preterm infants are no longer fed with human milk after the diagnosis of CMPA is made, which is in contrast with current nutrition guidelines in preterm infants. We strongly advocate that human milk with mothers on a cow's milk-free diet is the first choice of feed after the diagnosis of CMPA. Prospective studies are necessary to obtain more information regarding clinical presentation, diagnostic tools, and therapeutic approaches.

Introduction

Human milk provides the optimal nourishment for the preterm infant and is therefore the recommended feeding for preterm infants.^1,2 In addition, it needs to be considered as part of the therapeutic strategies to decrease preterm infants' mortality and morbidities. After all, the use of mother's own milk results in a significant lower incidence of necrotizing enterocolitis (NEC), sepsis, bronchopulmonary dysplasia (BPD), and retinopathy in prematurity.³ Human milk supports the development of the host defense of the preterm infant. Furthermore, clinical feeding tolerance is improved, and full enteral nutrition is obtained more quickly when fed with human milk. Finally, the neurodevelopmental outcomes are better when the preterm infant is on a predominantly human milk diet.¹

Due to the high protein need for fetal growth and to prevent extrauterine growth restriction, the use of human milk fortifier (HMF) that supplements human milk is essential in preterm infants. The most commonly used form of fortification is bovine-based milk fortifier. Human milk-derived fortification is now also available, although it is preliminary and expensive.^4,5

Nonetheless, providing human milk for a preterm infant can be a challenge sometimes.¹ When mother's own milk is unavailable or there is an insufficient supply of mother's own milk, then donor human milk or preterm formulas can be used as alternative options. A Cochrane review compared the use of donor human milk versus preterm formula for feeding preterm infants. It demonstrated that preterm formula was associated with higher rates of weight gain and growth. However, it also showed that preterm formula is associated with a higher risk of severe NEC.⁶ Therefore, several guidelines stress the importance of mother's own milk being superior to formula for preterm infants.^7,8

Cow's milk protein allergy (CMPA) is an immunological reaction against proteins found in cow's milk. Cow's milk contains multiple proteins such as casein and whey.⁹ CMPA is the most common food allergy in infancy, and can be IgE mediated or non-IgE mediated.^10,11 IgE-mediated CMPA gives an immediate allergic reaction, while the non-IgE pathway leads to a postponed response with mainly gastrointestinal symptoms.¹² CMPA and allergic proctocolitis at this age are predominantly non-IgE mediated.¹³ There are some diagnostic criteria for CMPA. First, elimination of cow's milk protein should provide resolution of symptoms. In addition, symptoms may recur if cow's milk protein is reintroduced into the diet. Finally, other causes for the symptoms must be ruled out.¹⁰

As opposed to term infants, the prevalence of CMPA in preterm infants is uncertain due to various reasons. First, CMPA has a wide range of symptoms, such as gastrointestinal (rectal blood loss, abdominal distension, vomiting, and feeding intolerance), dermatological, and respiratory (wheezing and respiratory distress).^10,14 The gastrointestinal symptoms of CMPA overlap with those of NEC. Therefore, it can be a challenge to distinguish CMPA from NEC. NEC is a life-threatening neonatal condition that causes inflammation of intestinal tissue.¹⁵

This inflammatory bowel disease is one of the most common gastrointestinal emergencies among preterm infants with significant morbidity and mortality.¹⁶ As a result of this, it is often the first working hypothesis for a neonatologist when a preterm infant presents with rectal bleeding, abdominal distension, or feeding intolerance.^17–19 Second, there are no significant diagnostic tools for CMPA in preterm infants. The diagnosis of CMPA is generally based on the response to elimination diets and the recurrence of the symptoms when cow's milk protein is reintroduced into the diet.^10,18,19

Another issue we encounter is that there are no extensively hydrolyzed milk formulae that are nutritionally adequate for preterm infants.¹⁷

The aim of this study is primarily to investigate the presentation of CMPA in preterm infants, and second to compare the different diagnostic and therapeutic options. In addition, a comparison will be made between CMPA in prematurity and CMPA in term infants.

Materials and Methods

Search method

We searched for the relevant literature using the medical databases PubMed, Web of Science, and the Cochrane Library. Articles published before the date of April 2021 were included. Our initial search terms were “cow's milk protein allergy,” “cow's milk allergy,” “preterm infants,” “premature infants,” “allergic proctocolitis,” “allergic enterocolitis,” “rectal bleeding,” “food protein-induced enterocolitis,” “breastfeeding,” and “preterm formula.” Based on titles and abstracts, we selected the relevant literature, and excluded articles that were irrelevant to the research question. Reference lists of the selected articles were checked to identify additional articles.

An overview of our search strategy is shown in Figure 1. Because of the sparse literature, no additional exclusion criteria were applied, and we expanded our research question to compare CMPA between preterm and term infants.

FIG. 1.

Overview of search strategy.

Post hoc analysis

We performed a post hoc analysis on the 25 case reports included in this study. We obtained a summary of the clinical manifestations, diagnostic tools, and management strategies of CMPA in preterm infants. Certain demographics were excluded in our post hoc analysis because too few case reports provided information regarding these demographics.

Results

Table 1 gives an overview of the studies on CMPA in preterm infants. Table 2 summarizes the case reports. Table 3 provides a post hoc analysis of the case reports.

Table 1.

Overview of Studies on Cow Milk Protein Allergy in Preterm Infants

	Morita et al.²⁷	Cordova et al.³⁰	Miyazawa et al.²⁸	Dupont et al.¹³	D'Netto et al.³³
Type of study	Retrospective study	Retrospective study	Retrospective multicenter survey	Retrospective study	Retrospective and prospective portion
Cases (n)	12	14	111	14	25
Patient characteristics
Median birth weight (g)	1,589 (662–2,460)	958 (500–1,320)	2,614 (406–3,936)	Unknown	1,200 (555–3,000)
Gestational age (weeks)	31 (25–36)	Unknown	36.9 (25.3–42.0)	31	29 (24–36)
Male/female ratio	7/5	Unknown	3/2	9/5	8/17
Family history of atopy, n (%)	8 (67)	Unknown	Unknown	Unknown	14/23 (61)
Feed before symptoms, n (%)
Human milk	3 (25)	2 (14)	18 (16)	0	2 (8)
(Preterm) formula	5 (42)	6 (43)	26 (23)	14 (100)	6 (24)
Human milk + preterm formula	4 (33)	4 (29)	60 (54)	0	0
Human milk and HMF	0	0	1 (1)	0	0
(Preterm) formula and HMF	0	0	0	0	4 (16)
Extensively hydrolyzed formula	0	1 (7.1)	0	0	13 (52)
Soy formula	0	1 (7.1)	0	0	0
Medium-chain triglycerides	0	0	2/111 (2)	0	0
Unknown	0	0	2/111 (2)	0	0
Onset of symptoms (days)	23 (4–49)	Unknown	7 (0–67)	42	77.7 (11–165)
Clinical events (n and/or %)
Vomiting	9 (75)	Most common	39	5 (35.7)	13 (52)
Bloody stools	8 (67)	5 (36)	36	6 (42.9)	12 (48)
Abdominal distension	1 (8.3)	Most common	26	0	8 (32)
Diarrhea	0	0	0	3 (21)	9 (36)
Less vigorous sucking	1 (8.3%)	0	18	0	0
Gastrointestinal bleeding	1 (8.3%)	0	0	0	0
Feeding intolerance	0	0	0	0	14 (56)
Irritability	0	0	0	0	13 (52)
Systemic symptoms	0	0	20.7	0	0
Respiratory symptoms	1 (apnea) (8.3%)	0	3.6	0	0
Cutaneous	0	0	7.4	0	0
Failure to thrive	0	0	18.9	0	15 (60)
Circulatory failure and shock	0	0	7.4	0	1 (4)
Diagnostic investigations
Eosinophils in blood/eosinophilia	1,100 (1,74.5–5,625)	Unknown	50% (def: >700/μL)	Unknown	3 (12%) >450 cells/mm³
sIgE positive (n/total or %)	9.0 (1/11)	Unknown	30	0	Unknown
Stool cytology (eosinophilia positive)	100 (9/9)	Unknown	75 (46/61)	Unknown	Unknown
Rectal milk challenge test positive	100 (5/5)	Unknown	Not performed	Unknown	Unknown
Oral food challenge performed	Not performed	Unknown	26% (25/98)	Unknown	Unknown
Positive biopsy	Not performed	Unknown	Not performed	Unknown	15 (60)
Atopy patch test positive	Not performed	Unknown	Not performed	70 (7/10)	Unknown
Feed after diagnosis (n and %)
Extensively hydrolyzed formula	Unknown	6 (43)	Unknown	0	15 (60)
Amino acid-based formula	Unknown	8 (57)	Unknown	14 (100)	10 (40)

Table 2.

Overview of Case Reports

	Manuyakorn et al.¹⁹	Manuyakorn et al.¹⁹	Manuyakorn et al.¹⁹	Aktaş et al.²²	Aktaş et al.²²	Aktaş et al.²²	Aktaş et al.²²
Patient characteristics		Monozygotic twin	Monozygotic twin			Triplet	Triplet
Birth weight (g)	980	1,070	1,050	1,950	2,500	1,300	1,215
Gestational age (weeks)	33	27	27	32	36	31	31
Gender	Male	Female	Female	Female	Male	Female	Female
Family history of atopy	Allergic rhinitis	Allergic rhinitis	Allergic rhinitis	Unknown	Unknown	Unknown	Unknown
Feed before symptoms	Preterm formula	Preterm formula	Preterm formula	Preterm formula and human milk	Human milk	Human milk	Human milk and human milk fortifier
Symptoms
Onset symptoms (DOL)	3	10,19,23	10	12	4	16	20
Type of symptoms	- Abdominal distension- Hypoperistalsis- Feeding intolerance- Poor weight gain	- Abdominal distension- Feeding intolerance - frequent apnea with oxygen desaturation	- Abdominal distension- Feeding intolerance - frequent apnea with oxygen desaturation	- Minimal abdominaldistension - Bloody stools- diaper rash	Bloody stools	Abdominal distension	Abdominal distension
Diagnostic investigations
Eosinophilia	16.59 × 10⁹/L	2.72 × 10⁹/L	2.01 × 10⁹/L	0.7%	22.5%	10%	6.5%
sIgE (kUA/L)	7.77	0.39	0.23	Unknown	Unknown	Unknown	Unknown
Abdominal X-ray	Dilation of the stomach with normal bowel gas pattern	Decreased bowel gas pattern	Decreased bowel gas pattern	Normal	Normal	Dilated loops	Dilated loops
FOBT	Unknown	Unknown	Unknown	Positive	Positive	Positive	Positive
Therapy
Feed after diagnosis	Extensively hydrolyzed formula	Extensively hydrolyzed formula	Extensively hydrolyzed formula	Amino acid-based formula	Amino acid-based formula	Amino acid-based formula	Amino acid-based formula
Challenge	Introduced after one week elimination with recurrence of former symptoms	Unknown	Unknown	Unknown	Unknown	Unknown	Unknown
Follow-up	Age of 1: challenge with maculopapular rash	Remains good weight, further unknown	Remains good weight, further unknown	Unknown	Unknown	Unknown	Unknown

	Aktaş et al.²²	Baldassarre et al.³⁴	Baldassarre et al.³⁴	Vlieghe et al.²⁹	Vlieghe et al.²⁹	Vlieghe et al.²⁹
Patient characteristics		Monozygotic twin	Monozygotic twin
Birth weight (g)	1,495	Unknown	Unknown	1,145	910	710
Gestational age (weeks)	31	31	31	28 weeks +6 days	27	25 weeks +4 days
Gender	Male	Male	Male	Unknown	Unknown	Unknown
Family history of atopy	Unknown	Unknown	Unknown	Unknown	Unknown	Unknown
Feed before symptoms	Human milk and human milk fortifier	Cow's milk-based formula	Cow's milk-based formula	Human milk and human milk fortifier	Human milk and human milk fortifier	Human milk and human milk fortifier
Symptoms
Onset symptoms (DOL)	14	1 month	Unknown	14, 30	41	20
Type of symptoms	Abdominal distension	- Abdominal distension- Bloody stools- Vomiting	Bloody stools	- Abdominal distension- Bloody stools- Increased apneas (only DOL 30)	Bloody stools	Persistent abdominal distension
Diagnostic investigations
Eosinophilia	20%	Positive in stool	Positive in stool	1,800–6,900/mm³16–40%	2,700/mm³16%	3,300/mm³24%
sIgE (kUA/L)	Unknown	Unknown	Unknown	Unknown	Unknown	Unknown
Abdominal X-ray	Dilated loops	Normal	Unknown	1. Small bowel distension2. Normal	Normal	Unknown
FOBT	Positive	Unknown	Unknown	Unknown	Unknown	Unknown
Therapy
Feed after diagnosis	Amino acid-based formula	Extensively hydrolyzed and probiotic formula	Extensively hydrolyzed and probiotic formula	Human milk and extensively hydrolyzed formula	Human milk and amino acid-based formula	Human milk and extensively hydrolyzed formula
Challenge	Unknown	Performed 6 months after discharge with a negative result	Performed 6 months after discharge with a negative result	Unknown	Unknown	Unknown
Follow-up	Unknown	Started on a regular diet after the negative challenge test without any problems	Started on a regular diet after the negative challenge test without any problems	Intestinal symptoms did not recur	Unknown	Unknown

	Nakasone et al.²⁰	Preto et al.²³	Swart and Ultee²⁴	Coviello et al.¹⁸	Coviello et al.¹⁸	Lu and Zhang³⁶
Patient characteristics				Twin	Twin
Birth weight (g)	506	1,030	1,340	965	1,340	Unknown
Gestational age (weeks)	24	28	30	30	30	28
Gender	Male	Female	Female	Female	Female	Male
Family history of atopy	Unknown	Unknown	Negative	Unknown
Feed before symptoms	Preterm formula and human milk fortifier DOL 29: infant formula + hmf	Human milk and preterm formula	Preterm formula with fortified human milk	Human milk with fortified human milk	Human milk with fortified human milk	Expressed breast milk and preterm formula
Symptoms
Onset symptoms (DOL)	29,36,46, 85	17,45	8,12,14	10, 35	10, 27	29, 46, 54
Type of symptoms	- Recurrent abdominal distension + ascites - respiratory deterioration	- Abdominal distension- Vomiting- Cardiorespiratory instability- bloody stools	Bloody stools	- Lethargy- Vomiting- Bloody stools- Abdominal distension	Two NEC-like episodes: - abdominal distension- Bloody stools- Feeding intolerance- Lethargy- Bradycardia episodes	Recurrent episodes of bloody stools
Diagnostic investigations
Eosinophilia	Unknown	49%	3.61/nL	Normal	17.7%	0.4–4.6–13.8
Sige (kua/L)	Unknown	3.13	0.37	Negative	0.30	Unknown
Abdominal X-ray	Dilation of intestinal gas and elevated diaphragm	Gaseous distension and intestinal wall edema, no pneumatosis	No signs of NEC or obstruction	Normal	DOL 10: dilated bowel loopsDOL 14: pre-NEC signsDOL 27: pneumatosis	DOL 29: mild thickening of small bowel, no pneumatosis
Fobt	Unknown	Unknown	Unknown	Unknown	Positive	Positive
Therapy
Feed after diagnosis	Donated breast milk and mother's own milk	Human milk (mother on elimination diet) and EHF	Amino acid-based formula	Amino acid-based formula	Amino acid-based formula	Extensively hydrolyzed formula
Challenge	Unknown	Unknown	Reintroduction of cow's milk at age of 1 year succeeded without any problems	Unknown	Unknown	60 days after discharge: change to expressed mother's milk with no symptoms
Follow-up	Unknown	Unknown	Normal development and growth	2 months follow-up showed a good condition	2 months follow-up showed a good clinical condition	No bloody stools recurred, breastfeeding

	Srinivasan et al.¹⁷	Tekgündüz et al.¹⁴	Faber et al.³⁹	Ferretti et al.¹¹	Becker⁴⁰	Nakamura et al.²¹
Patient characteristics
Birth weight (g)	965	950	2,570	2,350	1,370	1,400
Gestational age (weeks)	26	28	33	33 + 2	31	29 + 2
Gender	Male	Unknown	Male	Male	Female	Male
Family history of atopy	Unknown	Allergic asthma in mother	Family history of CMPI	Unknown	CMPA: 1 parent and 1 sibling	Both parents: allergic rhinitis
Feed before symptoms	Expressed breast milk + HMF	Cow's milk-based preterm formula	Cow's milk-based formula	Expressed breast milk + cow's milk-based formula	Human milk and human milk fortifier	Human milk and human milk fortifier
Symptoms
Onset symptoms (DOL)	17	60, 80	1 (4h)	2 (32h)	21	12
Type of symptoms	DOL 17- Abdominal distension- Hypoperistalsis- Vomiting- Lethargy- Episodes of bradycardia with oxygen deprivation DOL 28 and 61- Abdominal distension- Bloody stools	- Bloody stools- Respiratory distress withoxygen need– Vomiting	Bloody stools	bloody stools	bloody stools	DOL 12 and 21 aggravation of respiratory symptoms:- Apnea- Wheezing
Diagnostic investigations
Eosinophilia	DOL 17: 40%DOL 28: 13%DOL 61:14%	Unknown	Unknown	3.2 × 10⁹/L	Unknown	Positive
sIgE (kUA/L)	Unknown	Negative	Unknown	Unknown	Unknown	Negative
Abdominal X-ray	Pneumatosis	Unknown	Unknown	Normal	Unknown	Unknown
FOBT	Unknown	Unknown	Unknown	Unknown	Unknown	Unknown
Therapy
Feed after diagnosis	DOL 28–58: extensively hydrolyzed formulaDOL 61: amino acid-based formula	Extensively hydrolyzed formula	Amino acid-based formula	Amino acid-based formula	Human milk (mother on elimination diet) and amino acid-based formula	Human milk only (excluding HMF)
Challenge	Two times (DOL 28 and 61) reintroduction of EBM with recurrence of symptoms	Reintroduction of cow's milk formula 2 weeks later: all symptoms reoccurred	Unknown	Challenge was introduced after 10 months; on follow-up he was symptom free	Unknown	Performed at 18 months: negative
Follow-up	Unknown	Unknown	Unknown	3 years after presentation: no recurrence of symptoms	Unknown	At the age of 2: normal development

DOL, day of life; HMF, human milk fortifier.

Table 3.

Post Hoc Analysis of Case Reports

Gender	Percentage	Interval
Male	44% (11/25)
Female	40% (10/25)
Unknown	16% (4/25)
Median birth weight (gram)	1,180	(506–2,570)
Median age of birth	30 weeks PML	(24–36)
Time to first onset developing symptoms (DOL)	14	(1–80)
Type of symptoms
Bloody stools	64% (16/25)
Abdominal distension	60% (15/25)
Respiratory symptoms	32% (8/25)
Vomiting	20% (5/25)
Feeding intolerance	16% (4/25)
Lethargy	12% (3/25)
Hypoperistalsis	8% (2/25)
Diaper rash	4% (1/25)
Feed before diagnosis
Human milk and human milk fortifier	40% (10/25)
Preterm formula	28% (7/25)
Human milk and preterm formula	16% (4/25)
Preterm formula and human milk fortifier	8% (2/25)
Human milk	8% (2/25)
Feed after diagnosis
Amino acid-based formula	44% (11/25)
Extensively hydrolyzed formula	28% (7/25)
Human milk and hydrolyzed formula	12% (3/25)
Human milk and amino acid-based formula	8% (2/25)
Donated human milk and mother's own milk	4% (1/25)
Human milk	4% (1/25)
Challenge
Positive	12% (3/25)
Negative	24% (6/25)
Not performed	64% (16/25)

Although we aimed to search for guidelines about CMPA in preterm infants, the literature was heterogeneous, scarce, and mainly based on retrospective studies and case reports. Overall, the majority of preterm infants who developed CMPA were exposed to bovine-based milk proteins before the development of the symptoms (cow milk protein-based preterm formula or bovine milk-based fortifier). In the studies, the percentage receiving a form of preterm formula based on cow's milk protein ranged from 75 to 100. In the post hoc analysis of the case reports, this percentage was 92.

The most common clinical manifestations of CMPA in preterm infants were bloody stools, vomiting, and abdominal distension (Tables 1 and 3). However, in some cases, respiratory symptoms such as apnea episodes were more prominent than gastrointestinal symptoms.^{14,17,19–21} Tekgündüz et al. published a case report of a preterm infant with BPD. The patient initially presented with respiratory symptoms, but at day of life (DOL) 80 there were also bloody stools. CMPA was suspected and confirmed with an oral food challenge (OFC): gastrointestinal and respiratory symptoms reoccurred. Thus, it was doubtful if the respiratory symptoms were a result of the BPD and the CMPA diagnosis.¹⁴

Several studies published a case of CMPA that presented as NEC and was treated as (recurrent) NEC.^{17,19,22–24} Some of these mimicked NEC both clinically and on X-rays.^17,18,20 CMPA and NEC have a lot of symptoms in common such as grossly bloody stools, therefore differentiating them can be difficult.^17,22 It is not uncommon that CMPA engages in NEC mimicking and is therefore initially treated as NEC, which includes stopping all enteral feeding. Consequently, the symptoms disappear with this approach. However, when restarting enteral nutrition, all symptoms reoccur.¹⁷

In the larger studies on CMPA in preterm infants, not one single study had human milk as the choice of feed after diagnosis (Table 1). Of the 25 cases, only 7 (28%) retained human milk in their diet after diagnosis (Table 3).

Discussion

In this literature study, we aimed to evaluate the presentation of CMPA in preterm infants, and to investigate the different diagnostic and therapeutic options. Although our goal was to search for guidelines about CMPA in preterm infants, we found that there were no good qualitative studies on CMPA in preterm neonates, which is the reason we were obligated to work mainly with case reports in this study.

Nevertheless, it is noteworthy that the majority of the preterm infants who developed CMPA were exposed to bovine-based milk proteins before the development of the symptoms (cow milk protein-based preterm formula or bovine milk-based fortifier). These findings are in line with a clinical trial where term newborn infants at risk of atopy were randomized to receive either exclusive human milk or human milk and cow's milk-based preterm formula in the first three days of life. Infants who exclusively received human milk had a significant decreased risk of developing CMPA, suggesting that sensitization to cow's milk and food allergy, including CMPA and anaphylaxis, is primarily preventable by avoiding cow milk supplementation for at least the first 3 days of life.²⁵

In addition to the already cited benefits of human milk, this is yet another argument for attempting to feed a preterm infant exclusively with human milk. If mother's own milk does not meet the caloric and nutritional needs of the preterm infant, one may consider using exclusive HMF instead of bovine-derived fortifier. Ananthan et al. compared the use of HMF with bovine milk-derived fortifier (BMF) in a meta-analysis. They concluded that HMF was associated with a lower risk of NEC, but also with poorer weight gain.²⁶

The incidence of CMPA in preterm neonates remains uncertain due to the lack of large prospective studies and the fact that CMPA is still often unrecognized in preterm infants.¹⁹ Morita et al. cited an incidence of CMPA of 1.1% in their neonatal ward.²⁷ This percentage is higher than the 0.21% reported in a retrospective multicenter survey of Miyazawa et al.²⁸ In addition, we found that there is a wide variety in the time to first onset of symptoms (DOL 0–165).

According to Vlieghe et al. and Tekgündüz et al., several factors contribute to the sensitization to cow's milk protein allergens.^14,29 An immature immune system and various harmful factors to the intestinal epithelium (free radicals, antibiotics, food restrictions, intestinal ischemia) lead to an increased intestinal permeability. Due to this elevated permeability, there is a sensitization to potential allergic allergens, such as cow's milk protein allergen.^14,18,29,30 Various cow milk proteins, such as casein, β-lactoglobulin, lactoferrin, and bovine immunoglobin, can lead to this sensitization.⁹ The fact that possible allergic antigens can move across the placenta and reach the amniotic fluid, and that some premature babies with CMPA were exclusively breastfed, suggests that the initial sensitization can already develop in utero.^9,31,32

Apart from the fact that NEC and CMPA have several symptoms in common, NEC could play a sensitizing role in the pathogenesis of CMPA. During the healing phase of NEC, proinflammatory β-lactoglobulin, casein-specific interferon-γ, and interleukin-4 increase. Once fully recovered from NEC, there is a considerable decrease of the aforementioned values and an increase of transforming growth factor-β1 (TGF-β1) cells. According to Cordova et al., it is possible that the shift from a proinflammatory phase to an increase in TGF-β1 cells is significant for the tolerance of various food antigens. Hence, the failure of this particular switch could contribute to the pathophysiology of the development of CMPA.³⁰

Nonetheless, several studies offer a different view on the hypothesis of the pathophysiology. Morita et al. analyzed the differences in manifestations of CMPA between term and preterm infants. In their retrospective study, all preterm infants only established CMPA later than the age of 32 weeks, corrected for gestational age. Hence, their interpretation is that the immune system before 32 weeks is still too immature to establish an autoimmune response such as CMPA and that a certain immunological maturation is necessary to develop CMPA.²⁷ This is an interesting hypothesis, although we must remember that these observations were made in a small group.

In accordance with the guidelines, the diagnosis of CMPA is solely based on the OFC.^9,15,16 Nevertheless, we note that an OFC is only performed in the minority of cases. This is due to the fact that it poses a potential risk of deteriorating the condition of the preterm infant, especially if they had already suffered from another medical condition. Therefore, the diagnosis of CMPA in preterm infants is mainly based on the elimination of cow's milk protein from their diet, rather than by applying the OFC.^18,27,28

Twelve percent of the cases reported a positive OFC, 24% a negative OFC, and the remaining 64% did not perform a challenge. However, these percentages of a positive and negative OFC should be interpreted with necessary nuances. The positive challenges were performed 1–2 weeks after eliminating cow's milk protein in their diet, whereas the challenges with a negative result were only performed at the age of 6–18 months.

In various studies, other different diagnostic tools are used to support the diagnosis of CMPA in preterm neonates. Peripheral eosinophilia is occasionally found in preterm infants with CMPA, and is therefore often determined in various cases.^13,15,33,34 Manuyakorn et al. describe three cases of preterm infants with CMPA and persistent peripheral eosinophilia, with a normalization or considerable decrease of the eosinophil level in the blood after elimination of cow's milk protein.¹⁹ However, several studies indicate that eosinophilia is a common and nonspecific finding during the first months of life of preterm infants.^32,35 It is also linked to infection, NEC, and transfusion of packed cells.^19,23,29,36

This does not signify that we cannot incorporate this finding into our decision tree when we consider CMPA and observe eosinophilia, especially when we notice a decrease or normalization of the eosinophilic count after eliminating cow milk protein from the diet.

Dupont et al. performed a retrospective study to investigate the potential of a ready-to-use atopy patch test (APT) as a diagnostic instrument in the diagnosis of CMPA, particularly in the non-IgE–mediated responses. The ready-to-use CM-APT contains a small chamber with powdered defatted milk. Once applied, the powder will dissolve on the skin. The control-APT and CM-APT remain 48 hours on the back of the preterm infant, and the interpretation is done 24 hours later. Fourteen preterm infants were suspected with CMPA, and were therefore switched to an amino acid-based formula.

Four of the preterm infants showed no improvement after this change in formula and were put back on CMP-based formula. Of the remaining 10 preterm neonates, 7 infants had a positive APT. Hence, this study argues that APT can be a noninvasive tool in the diagnosis of CMPA.¹³ To conclude, CMPA in preterm infants remains a difficult and main clinical diagnosis. Diagnostic tools to support the diagnosis are scarce and studied to a limited extent. Hence, it is important to consider CMPA especially in case of a preterm with suggestive symptoms not causing critical illness, or in case of a preterm infant with recurrent NEC symptoms after the reintroduction of enteral nutrition.

Preterm infants have higher nutritional and caloric needs than term infants. It is often challenging to meet these needs. With CMPA, there are a lot of additional challenges since the bovine-based milk fortifier cannot be used.²⁴ As mentioned before, human milk is the best choice of nutrition for premature babies. Therefore, it is striking that human milk almost completely disappears from the nutritional options after CMPA is diagnosed in preterm infants. As mentioned in the results, there is not one study that chooses human milk after the diagnosis. Of the 25 cases, only 7 (28%) retained human milk in their diet after diagnosis (Table 3). This is contradictory to the current guidelines regarding the most nutritionally adequate feed for preterm infants.^10,18,19

With the disappearance of human milk in the feed after diagnosis, all the advantages of human milk for the preterm infant automatically disappear as well. We therefore strongly advocate that human milk with mothers on a cow's milk protein-free diet is the first choice of feed after the diagnosis of CMPA, and that further research is moving in that direction. Moreover, it remains doubtful whether extensively hydrolyzed formula (EHF) is nutritionally competent for preterm infants.

A Cochrane systematic review concluded that it requires larger and more pragmatic studies to determine whether EHF is effectively nutritionally (in)adequate as a feed for preterm infants.² Zuppa et al. investigated the nutritional efficacy of EHF in preterm infants through a randomized controlled trial. The preterm infants nourished with EHF showed a slower weight evolution than those fed with standard preterm formula.³⁷ The influence of marketing of commercial milk formula on health professionals and science should also not be underestimated in this light, as recently reported in The Lancet Breastfeeding series.³⁸

It was noticed that there were three pairs of monozygotic twins who all had CMPA (Table 2). It is known that food allergies have genetic and environmental risk factors.³² These findings support the fact that CMPA has a genetic constitution.

Morita et al. compared CMPA in term and preterm infants. Their findings were that term and preterm infants have similar types of symptoms. However, there was a significant difference in median age of onset; preterm infants develop these symptoms much later than term infants (23 versus 3.5 days). Based on these findings, their hypothesis of immunological maturation to develop CMPA was substantiated.²⁷

We are aware that this study has several limitations: we were obligated to work with small heterogeneous studies, which were mostly case reports, and therefore we could only perform a post hoc analysis that automatically inherits bias. Therefore, it is crucial to interpret the analysis with the necessary caution. This made it challenging to compare results and get an overview. However, we believe it is important that this study was conducted: relevant literature on this particular subject was obtained, and diagnostic and therapeutic choices were compared with the known guidelines. The remaining questions and problems were presented, so that directions for further research are clarified.

Therefore, it would be interesting to set up a study that examines the effect of an elimination diet (preferably mother's own milk with mothers on a cow's milk protein-free diet) and evaluate the response. In case of a clinical response, an OFC should be performed after a period of 6 months to get a certain diagnosis.

Conclusions

It is important to be aware that CMPA in preterm infants can occur, and consider the diagnosis of CMPA in preterm infants with gastrointestinal or respiratory symptoms. Because of the aforementioned benefits of human milk, we emphasize the importance of mother's own milk in nutritional policies in preterm infants, even after the diagnosis of CMPA. Prospective and larger homogeneous studies are necessary to investigate the different aspects of CMPA in preterm infants.

Footnotes

Authors' Contributions

A.E. developed the concept of the article and conducted revisions. M.F. conducted the literature review and wrote the article. M.F. and A.E. discussed about the content of the article. All authors read and approved the final article.

Ethical Approval and Consent to Participate

No human or animal applicants were used in this article. This article review does not require an ethical approval.

Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

Gartner

, Morton

, Lawrence

, et al. Breastfeeding and the use of human milk. Pediatrics, 2005; 115(2):496–506.

DHC

, Klassen

, Embleton

, et al. Protein hydrolysate versus standard formula for preterm infants. Cochrane Database Syst Rev, 2019; 7(7):Cd012412.

Pados

. State of the science on the benefits of human milk for hospitalized, vulnerable neonates. Nurs Womens Health, 2023; 27(2):121–140.

Jupe

, Maslin

. The use of breast milk fortifier in preterm infants by paediatric dietitians in the UK. J Hum Nutr Diet, 2021; 34(1):24–32.

Premkumar

, Pammi

, Suresh

. Human milk-derived fortifier versus bovine milk-derived fortifier for prevention of mortality and morbidity in preterm neonates. Cochrane Database Syst Rev, 2019; 2019(11):CD013145.

Quigley

, Embleton

, McGuire

. Formula versus donor breast milk for feeding preterm or low birth weight infants. Cochrane Database Syst Rev, 2019; 7(7):CD002971.

Boies

, Vaucher

. ABM Clinical Protocol #10: Breastfeeding the Late Preterm (34–36 6/7 Weeks of Gestation) and Early Term Infants (37–38 6/7 Weeks of Gestation), Second Revision 2016. Breastfeed Med, 2016; 11:494–500.

Lapillonne

, Bronsky

, Campoy

, et al. Feeding the late and moderately preterm infant: A position paper of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr, 2019; 69(2):259–270.

Burris

, Burris

, Järvinen

. Cow's milk protein allergy in term and preterm infants: Clinical manifestations, immunologic pathophysiology, and management strategies. Neoreviews, 2020; 21(12):e795–e808.

10.

Mousan

, Kamat

. Cow's milk protein allergy. Clin Pediatr, 2016; 55(11):1054–1063.

11.

Ferretti

, Pilon

, Boland

, et al. Early onset allergic proctitis in a preterm neonate: A case report and review of the literature. Pediatr Dev Pathol, 2019; 22(2):152–156.

12.

Meyer

, Chebar Lozinsky

, Fleischer

, et al. Diagnosis and management of Non-IgE gastrointestinal allergies in breastfed infants-An EAACI Position Paper. Allergy, 2020; 75(1):14–32.

13.

Dupont

, Soulaines

, Lapillonne

, et al. Atopy patch test for early diagnosis of cow's milk allergy in preterm infants. J Pediatr Gastroenterol Nutr, 2010; 50(4):463–464.

14.

Tekgündüz

, Demirelli

, Caner İ. Cow's milk allergy in preterm infant with bronchopulmonary dysplasia. Turk Thorac J. 2015; 16(3):151–153.

15.

Neu

, Walker

. Necrotizing enterocolitis. N Engl J Med, 2011; 364(3):255–264.

16.

Schnabl

, Van Aerde

, Thomson

, et al. Necrotizing enterocolitis: A multifactorial disease with no cure. World J Gastroenterol, 2008; 14(14):2142–2161.

17.

Srinivasan

, Brandler

, D'Souza

, et al. Allergic enterocolitis presenting as recurrent necrotizing enterocolitis in preterm neonates. J Perinatol, 2010; 30(6):431–433.

18.

Coviello

, Rodriquez

, Cecchi

, et al. Different clinical manifestation of cow's milk allergy in two preterm twins newborns. J Matern Fetal Neonatal Med, 2012; 25 Suppl 1:132–133.

19.

Manuyakorn

, Benjaponpitak

, Siripool

, et al. Cow milk protein allergy presenting as feeding intolerance and eosinophilia: Case reports of three preterm neonates. Paediatr Int Child Health, 2015; 35(4):337–341.

20.

Nakasone

, Fujioka

, Suga

, et al. A preterm case of cow's milk allergy presenting with recurrent ascites treated with donor breast milk. Int J Environ Res Public Health, 2021; 18(3):1187.

21.

Nakamura

, Hatanaka

, Kashima

, et al. A male preterm infant with cow's milk allergy to human milk fortifier showing only severe respiratory symptoms. Fukushima J Med Sci, 2019; 65(2):50–54.

22.

Aktaş

, Ergenekon

, Ünal

, et al. Different presentations of cow's milk protein allergy during neonatal period. Turk J Pediatr, 2017; 59(3):322–328.

23.

Preto

C OD

, Braga

, Lima

, et al. Allergic enterocolitis in a preterm newborn. J Pediatr Neonatal Individ Med, 2017; 6(1):e060116.

24.

Swart

, Ultee

. Rectal bleeding in a preterm infant as a symptom of allergic colitis. Eur J Pediatr, 2003; 162(1):55–56.

25.

Urashima

, Mezawa

, Okuyama

, et al. Primary prevention of cow's milk sensitization and food allergy by avoiding supplementation with cow's milk formula at birth: A randomized clinical trial. JAMA Pediatr, 2019; 173(12):1137–1145.

26.

Ananthan

, Balasubramanian

, Rao

, et al. Human milk-derived fortifiers compared with bovine milk-derived fortifiers in preterm infants: A systematic review and meta-analysis. Adv Nutr, 2020; 11(5):1325–1333.

27.

Morita

, Iwakura

, Ohtsuka

, et al. Milk allergy in the neonatal intensive care unit: Comparison between premature and full-term neonates. Asia Pac Allergy, 2013; 3(1):35–41.

28.

Miyazawa

, Itabashi

, Imai

. Retrospective multicenter survey on food-related symptoms suggestive of cow's milk allergy in NICU neonates. Allergol Int, 2013; 62(1):85–90.

29.

Vlieghe

, Des Roches

, Payot

, et al. Human milk fortifier in preterm babies: Source of cow's milk protein sensitization?. Allergy, 2009; 64(11):1690–1691.

30.

Cordova

, Sriram

, Patton

, et al. Manifestations of cow's-milk protein intolerance in preterm infants. J Pediatr Gastroenterol Nutr, 2016; 62(1):140–144.

31.

Alabsi

, Reschak

, Fustino

, et al. Neonatal eosinophilic gastroenteritis: Possible in utero sensitization to cow's milk protein. Neonatal Netw. 2013; 32(5):316–322.

32.

ABM Clinical Protocol #24: Allergic Proctocolitis in the Exclusively Breastfed Infant. Breastfeed Med, 2011; 6(6):435–440.

33.

D'Netto

, Herson

, Hussain

, et al. Allergic gastroenteropathy in preterm infants. J Pediatr, 2000; 137(4):480–486.

34.

Baldassarre

, Cappiello

, Laforgia

, et al. Allergic colitis in monozygotic preterm twins. Immunopharmacol Immunotoxicol, 2013; 35(1):198–201.

35.

Juul

, Haynes

, McPherson

. Evaluation of eosinophilia in hospitalized preterm infants. J Perinatol, 2005; 25(3):182–188.

36.

, Zhang

. Food protein-induced enterocolitis syndrome presenting after necrotizing enterocolitis in a preterm neonate: A case report. Transl Pediatr, 2021; 10(5):1393–1398.

37.

Zuppa

, Visintini

, Cota

, et al. Hydrolysed milk in preterm infants: An open problem. Acta Paediatr, 2005; 94(449):84–86.

38.

Rollins

, Piwoz

, Baker

, et al. Marketing of commercial milk formula: A system to capture parents, communities, science, and policy. Lancet, 2023; 401(10375):486–502.

39.

Faber

, Rieu

, Semmekrot

, et al. Allergic colitis presenting within the first hours of premature life. Acta Paediatr, 2005; 94(10):1514–1515.

40.

Becker

. Cow's milk allergy in a preterm infant: a case presentation. Infant Child Adolesc Nutr, 2009; 1(4):214–219.